Biodiesel, a mixture of fatty acid methyl/ethyl esters (“FAME”), derived from plant/animal triglycerides through transesterification with an alcohol, is a fuel that is under a great deal of consideration. It has been assessed that biodiesel yields 93% more energy than that invested in its production and, relative to fossil fuels, greenhouse gases are reduced 41% by the biodiesel production and combustion while less air pollutants are released per net energy gain (Hill et al., Proc. Natl. Acad. Sci. USA 103:11206-11210 (2006)). While worldwide triglyceride sources are diverse, over 90% of biodiesel in the United States is made from soybean oil (“SBO”) (Collins, K. Statement of Keith Collins, chief economist, U.S. Department of Agriculture before the U.S. Senate Committee on Appropriations Subcommittee on Agriculture, Rural Development, and Related Agencies. Aug. 26, 2006). By increasing the production of soybeans there could be a positive feedback on agriculture through higher quantity of soybean meals and implicitly more meat for food and more triglyceride supply for biodiesel production.
Although these benefits are very attractive, the current biodiesel final cost is prohibitively high without governmental subsidies. Much of the actual technological complexity, involving multiple steps on triglyceride pretreatment and biodiesel separation/purification, originates from contaminants in the feedstock (e.g., water and free fatty acids (“FFA”)) or impurities in the final product (e.g., glycerol, methanol and soaps) (Van Gerpen, Fuel Process. Technol. 86:1097-1107 (2005)). Compared with these conventional catalytic methods, a relatively new supercritical method was found capable to reduce transesterification time from hours to minutes through a continuous process which requires no feed pretreatment for triglyceride with high FFA/water content (Kusdiana et al., Bioresour. Technol. 91:289-295 (2004)).
However, many factors have been found to affect the FAME yield in the supercritical method, including transesterification temperature, pressure and residence time, alcohol to triglyceride ratios, feedstock composition, as well as mixing and solubility parameters. Among these factors, the most important are the ratio of alcohol to triglyceride and the transesterification temperature. The former was found to increase the FAME yield when it was far beyond the stoichiometric molar ratio of 3:1 (up to 64:1), but it was associated with increased cost of pumping, separating, and recycling of the excess alcohol. The latter led to shorter residence time but higher energy consumption and the risk of FAME decomposition.
To understand these difficulties, a thorough survey was performed on increasingly reported information regarding the supercritical methods of biodiesel production (Kusdiana et al., Bioresour. Technol. 91:289-295 (2004); Han et al., Process Biochem. 40:3148-3151 (2005); Iijima et al, ASAE/CSAE Annual International Meeting, Ottawa, Ontario, Canada, Aug. 1-4, 2004, Paper no. 046073; Saka et al., Fuel 80:225-231 (2001); Kusdiana et al., Fuel 80:693-698 (2001); Bunyakiat et al., Energy & Fuels 20:812-817 (2006); Busto et al., Energy & Fuels 20:2642-2647 (2006); Warabi et al., Bioresour. Technol. 91:283-287 (2004); Demirbas, Energy Convers. Manage. 43:2349-2356 (2002); Kusdiana et al., J. Chem. Eng. Jpn. 34:383-387 (2001); He et al., Fuel 86:442-447 (2007); Diasakou et al., Fuel 77 :1297-1302 (1998); Cao et al., Fuel 40:347-351 (2005); Madras et al., Fuel 83:2029-2033 (2004); Varma et al., Ind. Eng. Chem. Res. 46:1-6 (2007); Demirbas, Energy Convers. Manage. 44:2093-2109 (2003); Han et al., Process Biochem. 40:3148-3151 (2005); Tijima et al., ASAE/CSAE Annual International Meeting, Ottawa, Ontario, Canada, Aug. 1-4, 2004, Paper No. 046073). In one example, refined SBO was treated with supercritical methanol and CO2 as a co-solvent at 280° C. and 143 bar in a batch reactor for 10 minutes (Han et al., Process Biochem. 40:3148-3151 (2005)). Under these conditions, 98.5% conversion of oil to biodiesel was reported. The other key process variables were molar ratios of methanol to oil (24:1) and CO2 to methanol (1:10). The reaction products were settled for 60 minutes for glycerol separation and then methanol was evaporated from both phases at 70° C. Although CO2 lowered the pressure and temperature (“P-T”) conditions of transesterification, the tradeoff between the process time and transesterification temperature was not well balanced. Moreover, the excess alcohol precluded obtaining usable biodiesel directly from the reactor.
In another example, canola oil and supercritical methanol were preheated at 270° C. and the mixture was then treated in a capillary reactor up to 500° C. and 400 bar for 4 minutes (Iijima et al., ASAE/CSAE Annual International Meeting, Ottawa, Ontario, Canada, Aug. 1-4, 2004, Paper No. 046073). The range of methanol to oil ratio was from 11:1 to 45:1 on molar basis. Thermal decomposition of glycerol was reported for temperatures beyond 400° C. At these high P-T values, the unsaturated high-molecular FAME also decomposed to C6-C10 smaller molecular esters. The optimum reaction temperature was considered 450° C. The excess methanol, up to 60%, was removed from the reaction products by using a rotary evaporator. In this case, too, the excess alcohol precluded obtaining usable biodiesel directly from the reactor.
Information on mutual solubility of the reaction components, often an overlooked issue, is essential for the production design and process operation. The reactants (triglycerides and alcohol) and the products (FAME and glycerol) are partially mutually soluble in the transesterification process. The alcohol is soluble in both FAME and glycerol, but is not significantly soluble in oil. With an increase in the mass fraction of FAME, the alcohol solubility in the triglyceride-FAME phase increases. The transesterification reaction is carried out in the alcohol phase and, consequently, the reaction advance depends on oil solubility in this phase. For example, when FAME content increases to 70%, the triglyceride-methanol-FAME mixture becomes a homogeneous phase (Zhou et al., J. Chem. Eng. Data 51:1130-1135 (2006)). Glycerol has a low solubility in both oil and FAME but high affinity for alcohol.
Kinetic studies on triglyceride-alcohol systems revealed an unusual behavior of the reaction rate constant with increasing temperature and pressure. To explain this phenomenon, the phase equilibria of the pseudo binary system sunflower oil (“SFO”)-methanol were measured at different temperatures between 200 and 230° C. and pressures between 10 and 56 bar (Glisic et al., J. Serb. Chem. Soc. 72:13-27 (2007)). The reported data indicated a strong influence of the phase equilibrium on the reaction kinetics. High-pressure phase equilibria have also been calculated for the ternary system of C54 triglyceride-ethanol-CO2, at 40-80° C. and 60-120 bar (Geana et al., Supercrit. Fluids 8:107-118 (1995)). The role of CO2 as cosolvent in increasing the mutual solubility of oil and ethanol was revealed.
Also, fluid transport properties play an important role in the transesterification reactions. In an example, the influence of the axial dispersion on the performance of tubular reactors during non-catalytic supercritical transesterification of triglycerides has been studied (Busto et al., Energy & Fuels 20:2642-2647 (2006)). It was found that supercritical transesterification reactors must be operated at axial Peclet numbers higher than 1000 in order to limit back mixing effects and achieve batch-like conversions at short residence times. Otherwise, the authors concluded that high temperatures and high methanol to oil ratios were required for high conversions at lower Peclet numbers.
Overall, these reports present the major findings on theoretical and technical aspects of biodiesel production by supercritical methods. Given the high number of the process parameters affecting supercritical transesterification conversions, many of them being competitive, it is difficult to comprehensively and cohesively grasp their effects. Indeed, there are not even two reported sets of data with similar claimed optimum transesterification conditions (Table 1).
TABLE 1Conditions and Yield for the Reported Noncatalytic TransesterificationOil/cosolventT (° C.)P (bar)MeOH/TGτ (min)B/CYield (%)Ref.Rapeseed350450424B95 1Soybean235626-27600B85 2Soybean/CO2 *2801432410B98 3Soybean/C3H8 **2801282410B98 4Canola420-45040011-454C~100 5Coconut and palm350190427C95-96 6Soybean3103504025C77-96 7Sunflower3502004040B96 8Castor and linseed3502004040B98 9Soybean/C3H8 **288966410B9910Soybean285-290100-11010-12N/AB~10011Waste oil/C3H8 **2801282417C9512Soybean/CO2 *350-425100-2503-62-3C~100a* CO2/MeOH = 0.1;** C3H8/MeOH = 0.05;1 Saka et al., Fuel 80: 225-231 (2001); Kusdiana et al., Fuel 80: 693-698 (2001);2 Diasakou et al., Fuel 77: 1297-1302 (1998);3 Han et al., Process Biochem. 40: 3148-3151 (2005);4 Cao et al., Fuel 40: 347-351 (2005);5 Iijima et al., ASAE/CSAE Annual International Meeting, Ottawa, Ontario, Canada, Aug. 1-4, 2004, Paper No. 046073;6 Bunyakiat et al., Energy & Fuels 20: 812-817 (2006);7 He et al., Fuel 86: 442-447 (2007);8 Madras et al., Fuel 83: 2029-2033 (2004);9 Varma et al., Ind. Eng. Chem. Res. 46: 1-6 (2007);10 Hegel et al., Ind. Eng. Chem. Res. 46: 6360-6365 (2007);11 D'Ippolito et al., Energy & Fuels 21: 339-346 (2007);12 Kasteren et al., Resour., Conserv. Recycling 50: 442-458 (2007);a present application, limited number of experiments were executed with this cosolvent as shown in Table 2, infra.Also, technical and economic feasibility studies of creating an alternative to conventional biodiesel industry are scarce (D'Ippolito et al., Energy & Fuels 21:339-346 (2007); Kasteren et al., Resour. Conserv. Recycling 50:442-458 (2007)).
Two major issues on improving the efficiency of biodiesel production concern the oil-alcohol mixing and the separation/purification processes. While the former issue can be overcome by carrying out transesterification in supercritical states, the latter is more challenging. As biodiesel production has risen, the excessive supply of glycerol has glutted the market, sinking its price. The price of glycerol ($0.20-$0.50/lb) could drop further as biodiesel production increases (Rosner, New York Times, Aug. 8, 2007). It would be desirable, therefore, to identify process parameters that allow high yield while affording decomposition of glycerol (even with slight decomposition of FAME), which will overcome the problem of glycerol recovery in the costly separation/purification steps.
The present invention is directed to overcoming these and other limitations in the art.